Fuzzy comprehensive evaluation of economic performance on thermal power plant using modified grey clustering

摘要

Real-time comprehensive evaluation of economic performance on thermal power plant unit is helpful to improve benefits of power plant. Due to the complexity of power plant, this paper proposes a three-layer evaluation framework to solve the real-time economic evaluation problem. The top layer only has a comprehensive evaluation objective. The second layer is block layer with four economic blocks, in which each block is achieved by their corresponding indexes of the third layer. The index selection of the evaluation framework is completed with PCA. Moreover, this paper proposes a combined weight strategy which is based on AHP and entropy method to determine element weight of each layer. As for evaluation strategy, this paper proposes a modified grey clustering for block evaluation. Based on grey clustering results and combined weight, an improved fuzzy comprehensive evaluation method is used to complete economic comprehensive assessment. The real data from a 1000MW power plant is used to train and test the evaluation model. Evaluation results are basically accord with real condition, which proves the effectiveness of the evaluation method.